This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. With completion of our recent upgrades to the integrated 95 GHz spectrometer extended-interaction klystron (EIK) pulse modulator, detailed in Subproject 0252, ACERT has entered a new era of significantly improved high-power pulse HF ESR spectroscopy. Initial high-power r.f. tests in the complete spectrometer indicate that the EIK tube bias and gated beam "on" time can now be adjusted for essentially unmeasurable power droop for intervals up to approximately 800ns. It has become evident from our initial post-upgrade empirical adjustments and observations, however, that our EIK biasing requirements under high-power pulsing are somewhat different than the manufacturer's (CPI, Inc., Canada) specifications indicate. Therefore, we have initiated a modest program for achieving a more complete picture of tube performance under different bias and drive conditions. Prior to the recent upgrade program, we had found that the EIK exhibited significant power droop during a (at that time fixed) 1[unreadable]s beam "on" time. In saturated mode the droop was measured to be a tolerable value of 21% of initial output power but the telling measurement of unsaturated droop indicated an extremely high value of 78% power loss over the beam pulse interval. Following incorporation of the modulator upgrades and repair of a modulator focus electrode bias supply offset due to component failure, approximately the same droop values were observed. However, a dramatic reduction of the EIK power droop characteristic was observed with adjustment of the h.v. switch supply above the previous 3KV limit. This observation indicates that the required focus electrode potential for optimum r.f. power extraction during the beam "on" time may be more positive than the 0[unreadable]20V specification supplied by CPI. We will in the near term investigate the best immediate bias adjustment for simultaneously minimizing power droop and overvoltage risk the modulator's expensive h.v. switch. These studies will be performed over the normal range of operating parametric conditions. In the long-term we will remove the modulator box with the goal of determining necessary modifications to the modulator to eliminate the droop effect entirely.